WEBVTT - FORENSICS: The Cattai Double Murder 0:00:02.570 --> 0:00:04.250 A listener production. 0:00:09.090 --> 0:00:13.739 Welcome to Crime Insider's forensics. For those joining us for 0:00:13.740 --> 0:00:17.820 the first time. My name's Catherine Fox. I'm a former GP, 0:00:18.000 --> 0:00:23.400 crime author and screenwriter. I'm enthralled by forensics and have 0:00:23.400 --> 0:00:27.720 spent thousands of hours researching for books and screenplays. So 0:00:27.720 --> 0:00:32.340 I thought, why not turn my research into a podcast? 0:00:32.909 --> 0:00:36.810 Every week you'll be joining me in discovering how forensic 0:00:36.810 --> 0:00:40.740 science is helping solve high profile crimes in Australia and 0:00:40.740 --> 0:00:46.920 around the world. This week how CCTV footage can lead 0:00:46.920 --> 0:00:51.360 to convictions, but also miscarriages of justice. 0:00:51.690 --> 0:00:54.480 Unless you can prove it scientifically, you can't really make 0:00:54.480 --> 0:00:57.450 that claim. And that was the fallacy of individualization. If 0:00:57.450 --> 0:00:59.520 you can't prove it scientifically, don't use it. 0:00:59.970 --> 0:01:03.060 Professor Glenn Porter is an expert in the field of 0:01:03.060 --> 0:01:08.550 forensic photography. He spent decades taking photos, analyzing them and 0:01:08.550 --> 0:01:10.830 presenting them as evidence in court. 0:01:10.830 --> 0:01:14.940 So it's very much a filmic visual narrative that crime 0:01:14.940 --> 0:01:16.920 scene operatives work on. 0:01:17.370 --> 0:01:20.520 It's easy to think that video and photographic evidence is 0:01:20.520 --> 0:01:23.699 the smoking gun and the slam dunk for a conviction, 0:01:25.110 --> 0:01:28.470 but it's actually far more complicated and not as clear 0:01:28.470 --> 0:01:32.400 cut as you might think. You'll hear from Glenn about 0:01:32.400 --> 0:01:36.780 how photos can be extremely helpful, but also can sometimes 0:01:36.780 --> 0:01:41.770 hinder a case for the prosecution. Glenn's taking us back 0:01:41.770 --> 0:01:44.170 to a high profile double murder. 0:01:46.940 --> 0:01:51.950 Peter Johnson was accused of murdering a couple at Cadi. 0:01:52.490 --> 0:01:57.380 The police facts allege that Mr. Johnson tortured these two 0:01:57.380 --> 0:02:00.770 people to get their Pin number, and then went around 0:02:00.770 --> 0:02:04.790 ATMs extracting money. Now, there was an ATM in Windsor, 0:02:05.150 --> 0:02:11.030 in western New South Wales, where the CCTV captured what 0:02:11.030 --> 0:02:14.810 police allege. Mr. Johnson taking the money out. But the 0:02:14.810 --> 0:02:18.080 quality of the CCTV was quite poor and you could 0:02:18.080 --> 0:02:21.020 see by the footage that Mr. Johnson was aware that 0:02:21.020 --> 0:02:23.330 there was a camera there. He did what I would 0:02:23.330 --> 0:02:26.660 call a recce. He walked past and noticed that the 0:02:26.660 --> 0:02:29.329 camera was there. So when he was extracting the money, 0:02:29.360 --> 0:02:32.450 he held up what looked like a white handkerchief or 0:02:32.450 --> 0:02:34.940 something like that over his face, so his face couldn't 0:02:34.940 --> 0:02:39.260 be identified. So police had the details of the ATM, 0:02:39.260 --> 0:02:43.550 obviously the account of the deceased, um, there was a 0:02:43.550 --> 0:02:48.800 slight time difference between the ATM and the CCTV timestamp, 0:02:48.800 --> 0:02:51.890 but that's that's fairly normal. The technology never lines up 0:02:51.889 --> 0:02:54.980 all the time, but the couple were dead in their 0:02:54.980 --> 0:02:58.250 home at that time that the money was extracted. You 0:02:58.250 --> 0:03:03.320 saw this person extracting the money. The police went and 0:03:03.320 --> 0:03:09.889 interviewed two people, his ex-wife and a friend of Mr. Johnson, 0:03:10.220 --> 0:03:14.150 and showed them the CCTV footage, which they were able 0:03:14.150 --> 0:03:18.800 to then identify Mr. Johnson through that video footage or 0:03:18.800 --> 0:03:22.369 the CCTV footage. When it came to Mr. Johnson's trial, 0:03:22.370 --> 0:03:26.540 I gave evidence for for the defence on the issues 0:03:26.540 --> 0:03:30.200 that may occur because in both of the statements and 0:03:30.200 --> 0:03:34.340 their oral evidence, they said they couldn't identify Mr. Johnson 0:03:34.340 --> 0:03:36.770 from his face because he was covering his face and 0:03:36.770 --> 0:03:41.330 it was fairly grainy, fairly poor quality. But they could 0:03:41.330 --> 0:03:45.350 recognise or identify Mr. Johnson through the way he walked. 0:03:45.680 --> 0:03:50.390 Now the problem I had with that, from an evidential perspective, 0:03:50.390 --> 0:03:54.680 I'm I'm not doubting what they claim, but the issue 0:03:54.680 --> 0:03:58.970 there from a technical perspective, was that the CCTV was 0:03:58.970 --> 0:04:02.210 shot around two frames a second, which results in a 0:04:02.210 --> 0:04:06.860 very jagged, animated type of motion. It's certainly not a 0:04:06.860 --> 0:04:10.100 natural motion, which is around 30 frames per second. So 0:04:10.100 --> 0:04:14.060 I was tasked to explain to the court how this 0:04:14.060 --> 0:04:18.020 could be misleading. The motion isn't a natural motion, so 0:04:18.020 --> 0:04:23.360 to make some form of recognition or identification could be dangerous. 0:04:23.720 --> 0:04:27.530 I gave my evidence and Judge Anthony Wheely very good. 0:04:27.770 --> 0:04:30.770 Supreme Court judge asked me a question. He said. So, 0:04:30.770 --> 0:04:35.089 doctor Porter, are you referring to a situation similar to 0:04:35.089 --> 0:04:36.320 a Charlie Chaplin film? 0:04:36.320 --> 0:04:37.760 The silent movies. 0:04:37.760 --> 0:04:41.810 The silent movies where the movement is quite animated, jerky 0:04:41.810 --> 0:04:43.310 and not natural? 0:04:43.490 --> 0:04:47.360 Well, that's probably because back then the silent films were 0:04:47.360 --> 0:04:52.070 recorded at about 16 to 18 frames per second. And 0:04:52.070 --> 0:04:54.470 I think for people to understand, it's almost like one 0:04:54.470 --> 0:04:57.500 of those flip books. So we used to draw as kids, 0:04:57.500 --> 0:04:59.600 you know, a whole thing of papers in a, in 0:04:59.600 --> 0:05:01.490 a book, in a corner, and then you can get 0:05:01.490 --> 0:05:05.299 your own little cartoon going. So it's easy to then 0:05:05.300 --> 0:05:09.350 understand in that context why this gentleman's gait would have 0:05:09.350 --> 0:05:11.240 appeared quite odd. 0:05:11.540 --> 0:05:14.330 And that distortion, I don't know how much it would 0:05:14.330 --> 0:05:18.080 affect their ability to make that judgment. That's that's not 0:05:18.080 --> 0:05:21.650 my call. But what was interesting, too, is, uh, Judge 0:05:21.650 --> 0:05:25.339 Whaley also instructed the jury, um, as a result of 0:05:25.339 --> 0:05:30.860 another Supreme Court matter, uh, Tang, he instructed the jury 0:05:30.860 --> 0:05:34.280 that the evidence given by the two witnesses is not 0:05:34.279 --> 0:05:39.350 identification evidence, but recognition evidence that they recognize Mr. Johnson 0:05:39.350 --> 0:05:40.280 on the video. 0:05:40.370 --> 0:05:45.020 The difference between thinking they recognize and identifying correctly. There 0:05:45.020 --> 0:05:47.990 seems to be a very big gap between the two. 0:05:48.020 --> 0:05:49.130 Just as a layperson. 0:05:49.580 --> 0:05:53.180 Yes. And there's a couple of distinctions. I mean, in 0:05:53.450 --> 0:05:58.969 facial identification, there's a distinction between facial recognition software, which 0:05:58.970 --> 0:06:02.539 enables a search of a database, and it gets to 0:06:02.540 --> 0:06:05.540 a what they call a recognized level. But to go 0:06:05.540 --> 0:06:10.310 the step to identification, it needs a forensic expert that's 0:06:10.310 --> 0:06:13.489 trained in facial identification to be able to make that 0:06:13.490 --> 0:06:16.099 claim a little bit like a fingerprint. They were used 0:06:16.100 --> 0:06:21.380 the automatic fingerprint identification system, Aphis or Nifrs, and to 0:06:21.380 --> 0:06:23.210 do a search of the fingerprint that they might have 0:06:23.210 --> 0:06:27.140 found in the crime scene, the AI or the computer 0:06:27.140 --> 0:06:32.000 technology can do a search of thousands in the database, 0:06:32.000 --> 0:06:36.049 but it's actually the forensic examiner that actually takes it 0:06:36.050 --> 0:06:40.550 to the identification level. And that's that's a system similar 0:06:40.550 --> 0:06:44.659 to forensic imaging. And there's been quite a bit of 0:06:44.660 --> 0:06:49.610 work done from. Universities around the world, in psychology departments, 0:06:49.610 --> 0:06:54.680 around people's ability to recognize people in images, and even 0:06:54.680 --> 0:06:59.630 in poor quality CCTV or poor quality video. And the 0:06:59.630 --> 0:07:03.710 results are quite consistent. The psychologists split it up into 0:07:03.710 --> 0:07:07.969 two categories familiar and unfamiliar faces. So if you know 0:07:07.970 --> 0:07:13.970 somebody even with poor quality CCTV, the accuracy is quite high. 0:07:14.150 --> 0:07:18.680 But if it's an unknown, an unfamiliar face, the error 0:07:18.680 --> 0:07:23.150 rate goes up quite significantly, up around 3,040%. So it's 0:07:23.150 --> 0:07:27.170 an interesting part of our innate ability to recognize people 0:07:27.170 --> 0:07:30.230 when we see them. Familiar faces. We do a pretty 0:07:30.230 --> 0:07:33.650 good job at recognizing them in images, even if they're 0:07:33.650 --> 0:07:34.400 poor quality. 0:07:35.450 --> 0:07:38.180 One of the interesting things, though, is comparing to a fingerprint. 0:07:38.180 --> 0:07:42.140 There's only a finite number of whorls, loops, things that 0:07:42.140 --> 0:07:45.950 they can find in combination. Surely in a human face 0:07:45.950 --> 0:07:54.380 there are so many variables length, distance, um, depth of muscle, fat, tissue. 0:07:54.380 --> 0:07:59.420 And then you have possible cosmetic interventions as well. Is 0:07:59.420 --> 0:08:02.660 a face like as unique as a fingerprint? 0:08:03.140 --> 0:08:07.160 Well, the difference major difference between face and fingerprint is 0:08:07.160 --> 0:08:12.020 that faces change. Fingerprints don't. So weight gain, weight loss, 0:08:12.020 --> 0:08:15.650 cosmetic surgery as you mentioned, age. That's one of the 0:08:15.650 --> 0:08:20.390 things that's made facial identification quite difficult that it's not consistent. 0:08:20.390 --> 0:08:24.260 You need a known source and exemplar to make that comparison. 0:08:24.260 --> 0:08:28.130 And that exemplar is obviously done sometime in the past. 0:08:28.130 --> 0:08:32.030 So depending upon how far in the past it was 0:08:32.420 --> 0:08:37.309 may affect the the accuracy. Martin Evison in 2014 wrote 0:08:37.309 --> 0:08:41.540 a brilliant paper called The Third Forensics, and he suggested 0:08:41.540 --> 0:08:46.760 that forensic identification is entering its third phase. So fingerprints 0:08:46.760 --> 0:08:51.410 being the first, DNA being the second method, and faces 0:08:51.410 --> 0:08:55.430 identification was going to be the third forensic identifier, but 0:08:55.429 --> 0:08:57.920 it just hasn't played out that that way. 0:08:58.309 --> 0:09:03.579 You mentioned the database with fingerprints. Facial databases. Do they 0:09:03.580 --> 0:09:07.630 help you in terms of helping to identify a person, 0:09:07.630 --> 0:09:12.250 in terms of the likelihood that these facial features are 0:09:12.250 --> 0:09:16.290 consistent with this? Person in this photograph. 0:09:16.920 --> 0:09:20.730 Yeah. So there are some systems out there that the 0:09:20.730 --> 0:09:24.839 Australian government are now using for facial identification. And we 0:09:24.840 --> 0:09:27.360 see it now on our mobile phones when we sign 0:09:27.450 --> 0:09:32.340 deeds and contracts, you can there, there are verification sort 0:09:32.340 --> 0:09:35.460 of processes that are using the face as a, as 0:09:35.460 --> 0:09:39.810 a form of verification based on your passport photographs or 0:09:39.809 --> 0:09:44.220 your license photograph. So that technology certainly coming to play. 0:09:44.220 --> 0:09:45.990 That's facial recognition isn't it. 0:09:45.990 --> 0:09:49.559 Facial recognition. Yeah. Not identification from a forensic perspective. Yeah. 0:09:49.559 --> 0:09:53.610 But but certainly the facial recognition is is playing a 0:09:53.610 --> 0:09:57.360 part with facial databases because we, we've been in New 0:09:57.360 --> 0:10:01.410 South Wales and other states, we've been collecting biometric data 0:10:01.410 --> 0:10:05.640 from our driver's licenses for 15, 20 years now. It's 0:10:05.640 --> 0:10:08.820 been quite a while since photo ID has been used 0:10:08.820 --> 0:10:13.410 in our driver's licenses. Those authorities, government authorities have been 0:10:13.410 --> 0:10:16.469 collecting that. Same with our passport. Our passport now has 0:10:16.470 --> 0:10:21.000 biometric information. A passport photograph has biometric information. So the 0:10:21.000 --> 0:10:24.990 state actually has as a resource quite a lot of 0:10:24.990 --> 0:10:27.930 people's faces in databases. How is the state going to 0:10:27.929 --> 0:10:31.260 use that. They've got some legislation at the moment that 0:10:31.530 --> 0:10:36.240 allows organisations to use it in those software applications. I 0:10:36.240 --> 0:10:39.630 just had a one of my PhD students who finished 0:10:39.630 --> 0:10:45.060 a research project looking at whether the facial databases and 0:10:45.059 --> 0:10:49.619 the facial recognition software, when you get poor quality CCTV 0:10:49.620 --> 0:10:53.970 where you can't really see facial detail. She tested whether 0:10:53.970 --> 0:10:59.130 forensic artists could be a transition between the CCTV of 0:10:59.130 --> 0:11:03.059 poor quality and the facial recognition software. So an artist 0:11:03.059 --> 0:11:07.950 sketch off the images and then the facial database can 0:11:07.950 --> 0:11:12.120 search the actual line drawing or the sketch, and she 0:11:12.120 --> 0:11:15.510 come up with some very interesting results that said that, yes, 0:11:15.510 --> 0:11:19.740 the technology can recognize, um, drawings. And in some aspects 0:11:19.740 --> 0:11:23.610 they were better than poor quality CCTV photographs. The artists 0:11:23.610 --> 0:11:27.929 are outperformed in in some areas, particularly areas of really 0:11:27.929 --> 0:11:31.170 bad angles when there was a sharp angle, not a 0:11:31.170 --> 0:11:35.339 normal frontal view, which is, um, straight on which I'm 0:11:35.340 --> 0:11:37.590 looking at the camera. Now, that's a normal frontal view, 0:11:37.590 --> 0:11:42.060 but any side angles or uh, high angles, which CCTV 0:11:42.059 --> 0:11:47.100 often is high, um, the, the artists actually outperformed in 0:11:47.100 --> 0:11:50.580 some instances than the photograph. So that was an interesting 0:11:50.580 --> 0:11:54.930 experiment that Vanessa, uh, gained a PhD out of. 0:11:55.200 --> 0:11:57.660 In terms of what can people do if there are 0:11:57.660 --> 0:12:00.090 so many images out there? And I just think how 0:12:00.090 --> 0:12:04.350 many selfies are being taken by criminals and non criminals 0:12:04.350 --> 0:12:08.130 every single day. There's got to be millions and millions 0:12:08.130 --> 0:12:13.140 of of photographs taken and uploaded. So potentially there's a 0:12:13.140 --> 0:12:17.940 massive database, a private companies utilizing that database. 0:12:19.020 --> 0:12:23.220 Yeah, it's a bit. It's. There is some legislation around 0:12:23.220 --> 0:12:28.350 where private enterprise can use software from the federal government, 0:12:28.620 --> 0:12:32.460 and they can access certain elements, but they don't access 0:12:32.460 --> 0:12:36.840 the database, but they can apply for a recognition search 0:12:36.840 --> 0:12:40.829 from different faces. I'm not familiar with the 100% with 0:12:40.830 --> 0:12:43.110 the legislation, but this is something that's come over the 0:12:43.110 --> 0:12:46.290 last couple of years. It's going to be obviously more 0:12:46.290 --> 0:12:50.070 prominent as we move into the AI tool of technology, 0:12:50.070 --> 0:12:53.550 I would imagine. I also had an honor student who's 0:12:53.550 --> 0:12:57.000 just finished an ex police officer, Tony Caledon, and his 0:12:57.000 --> 0:13:00.089 career was working in the police, in the BCI and 0:13:00.090 --> 0:13:05.850 also in intelligence and then privately over in private security 0:13:05.850 --> 0:13:08.099 before he retired. So he's got a sort of a 0:13:08.100 --> 0:13:10.920 background on both of those. And what he found in 0:13:10.920 --> 0:13:14.699 his research is these notion of fusion centers going to 0:13:14.700 --> 0:13:19.290 occur where surveillance of the of areas are going to 0:13:19.290 --> 0:13:24.480 be black box. So with AI being able to interpret 0:13:24.480 --> 0:13:28.740 activity rather than if you can imagine a security officer 0:13:28.740 --> 0:13:32.250 sitting in front of, you know, 200 screens, the the 0:13:32.250 --> 0:13:35.819 amount of detail becomes kind of white noise and detecting 0:13:35.820 --> 0:13:38.940 things that are happening is not very easily. Well, in 0:13:38.940 --> 0:13:42.890 fusion centers, they're all blacked out. Sort of uses a simple, 0:13:42.890 --> 0:13:48.530 oversimplified analogy. They're all blacked out until the AI detects 0:13:48.530 --> 0:13:53.929 some suspicious behavior, then it becomes visible, and then the 0:13:53.929 --> 0:13:58.460 security agency can then may make a decision on the action. 0:13:58.700 --> 0:14:03.020 So it's an interesting concept that Tony kind of discovered 0:14:03.020 --> 0:14:06.440 that with fusion centers, you're going to get more surveillance 0:14:06.440 --> 0:14:10.969 of the civilians or of the citizens, but more privacy, 0:14:11.300 --> 0:14:14.960 which is a bit, you know, sounds contradictory, but, um, 0:14:14.960 --> 0:14:17.689 there's some protection of the privacy because it's all black 0:14:17.690 --> 0:14:21.290 box until you see someone in a car park looking 0:14:21.290 --> 0:14:24.020 into car windows to see if there's anything you know 0:14:24.020 --> 0:14:26.630 in there to steal the. I would pick that up 0:14:26.630 --> 0:14:30.230 as suspicious activity and then turn on and alert the 0:14:30.230 --> 0:14:33.590 security agency to. Then they can make a decision on 0:14:33.590 --> 0:14:36.320 what that action is. So the technology is going to 0:14:36.320 --> 0:14:39.980 get very clever. Um, but I think it's good in 0:14:39.980 --> 0:14:43.340 one way that there thinking about the privacy because we 0:14:43.340 --> 0:14:46.970 don't want an over surveillance state. I don't think most 0:14:46.970 --> 0:14:50.840 people would want that. But this fusion center technology may 0:14:50.840 --> 0:14:55.610 have our interest in in heart, I guess, around privacy. 0:14:55.610 --> 0:14:57.620 It'd be interesting to see how it pans out over 0:14:57.620 --> 0:14:59.270 the next ten, 15 years. 0:14:59.480 --> 0:15:04.100 Well, we've also seen deepfake videos, and this is obviously 0:15:04.100 --> 0:15:07.670 a very divisive thing that you can, um, if you 0:15:07.670 --> 0:15:12.940 have the technology and the ability you can. Completely alter 0:15:12.940 --> 0:15:16.900 someone's face in a video and replace it with someone else's. 0:15:17.140 --> 0:15:20.440 For example, at the recent international forensic conference, one of 0:15:20.440 --> 0:15:26.670 the forensic scientists had Obama, President Obama. Actually introducing her, 0:15:26.910 --> 0:15:30.300 and everyone initially was like, oh, and then 11 or 0:15:30.300 --> 0:15:34.080 12 seconds into it, the mouth didn't sync with the dialogue, 0:15:34.080 --> 0:15:36.510 and so we all realized that it was a deep fake. 0:15:36.510 --> 0:15:40.110 But just initially there was a room full of forensic 0:15:40.110 --> 0:15:42.900 experts who were, oh my gosh, is that actually a 0:15:42.900 --> 0:15:46.920 bummer introducing her? And that was the whole point. If 0:15:46.920 --> 0:15:51.300 you could even convince a roomful of one of 1500 0:15:51.510 --> 0:15:56.820 forensic experts how easy it is to fool the public. 0:15:57.180 --> 0:16:00.780 That is very, very worrying. And, you know, can you 0:16:00.780 --> 0:16:03.990 trust anything in imaging? That's that's where we might go. 0:16:03.990 --> 0:16:08.880 But the filmmaker, Errol Morris, uh, has published a book 0:16:08.880 --> 0:16:12.210 called Believing Is Seeing. And I've got a quote from 0:16:12.210 --> 0:16:16.860 Morris book. It says photographs provide evidence, but no shortcut 0:16:16.860 --> 0:16:20.850 to reality. It is often said that seeing is believing. 0:16:20.850 --> 0:16:24.210 But what we do not form our beliefs on the 0:16:24.210 --> 0:16:28.290 basis of what we see. Rather, what we see is 0:16:28.290 --> 0:16:31.800 often determined by our beliefs. Believing is seeing, not the 0:16:31.800 --> 0:16:35.640 other way around. It sort of highlights the ambiguity sometimes 0:16:35.640 --> 0:16:39.810 of imaging as evidence. It's what we want to, uh, 0:16:40.050 --> 0:16:43.950 believing is seeing the imagery, the fake, the deep fake, 0:16:43.950 --> 0:16:47.310 or you know, that it's Obama and you know that 0:16:47.310 --> 0:16:51.000 he's talking. So you often would come to the belief 0:16:51.000 --> 0:16:53.010 that that's Obama talking. 0:16:54.280 --> 0:16:57.970 In terms of having a suspect in custody. We all 0:16:57.970 --> 0:17:02.200 know about mug shots. Profiles front on and each profile 0:17:02.200 --> 0:17:05.050 can vary too, so I gather they take each side. 0:17:06.330 --> 0:17:10.470 Do the police have a right to actually try and 0:17:10.470 --> 0:17:15.330 photograph from the same perspective, for example, as the ATM captured, 0:17:15.330 --> 0:17:21.210 to try and then superimpose the accused face onto the 0:17:21.210 --> 0:17:23.850 evidence that they have and see if that matches. 0:17:24.680 --> 0:17:29.990 We've seen this in some cases where anatomists have superimposed 0:17:29.990 --> 0:17:40.609 faces onto forensically obtain images. It's it's dangerous. Um. Superimposition can. 0:17:41.480 --> 0:17:45.530 Hide something as well as show something. So again, it 0:17:45.530 --> 0:17:49.970 can really superimposition can really trick people's brains. So it's 0:17:49.970 --> 0:17:53.060 I'm not much of a fan of it. This technology 0:17:53.060 --> 0:17:57.290 came out with identification of, of skeletal remains. So they 0:17:57.290 --> 0:18:00.679 would get a skull and get a photograph of the 0:18:00.680 --> 0:18:03.290 person or the deceased who they think it is, and 0:18:03.290 --> 0:18:08.780 then they would superimpose that image onto the skeletal, onto 0:18:08.780 --> 0:18:12.649 the skull, and match up certain anatomical points of the 0:18:12.650 --> 0:18:15.500 eye sockets and the teeth and, and those type of 0:18:15.500 --> 0:18:18.110 things to try to identify the body. And this is 0:18:18.109 --> 0:18:22.700 a technique designed before DNA. So DNA is obviously the 0:18:22.700 --> 0:18:25.760 the premium method for identification now. But if there's no 0:18:25.760 --> 0:18:29.570 DNA to match it to or no maternal DNA with 0:18:29.570 --> 0:18:33.919 mitochondrial DNA, this technique I guess, is still valid. If 0:18:33.920 --> 0:18:36.439 you talk to a fingerprint expert. What they don't do 0:18:36.440 --> 0:18:40.670 is superimposition, fingerprints, and there's a good reason for it 0:18:40.670 --> 0:18:41.480 because of the. 0:18:41.480 --> 0:18:43.850 Distortion they do on telly all the time. 0:18:44.750 --> 0:18:47.869 Yeah, they do a lot of things on tele. Um, 0:18:47.869 --> 0:18:51.919 it can be more misleading than give information so side 0:18:51.920 --> 0:18:57.200 by side, like for like combination is by a forensic 0:18:57.200 --> 0:19:01.609 expert is the method of choice rather than superimposition. It's 0:19:01.609 --> 0:19:06.590 too difficult. But image perspective which can change the face 0:19:06.590 --> 0:19:11.600 quite a lot. Um and image perspective is. The spice, 0:19:11.600 --> 0:19:16.280 how three dimensional objects are now recorded into two dimensional. 0:19:16.280 --> 0:19:19.460 And basically, if you got and you'll see this on 0:19:19.460 --> 0:19:22.820 a lot of selfies, you know, close up any close 0:19:22.820 --> 0:19:29.510 up photographs of people's faces. The image perspective is quite deep. 0:19:29.510 --> 0:19:32.240 So you might notice that, you know, the person's nose 0:19:32.240 --> 0:19:36.860 seems exaggerated in a very close up photograph. The further 0:19:36.859 --> 0:19:42.290 you move away, the flatter the perspective comes. And I've 0:19:42.290 --> 0:19:46.550 measured about four meters. After four meters it becomes fairly stable. 0:19:46.820 --> 0:19:50.750 So fashion photographers use this quite a lot as well. 0:19:50.750 --> 0:19:53.209 This technique, in the sense that, you know, when you're 0:19:53.210 --> 0:19:58.520 using people that have this beautiful proportionality of their face 0:19:58.520 --> 0:20:01.040 and that's what makes them pretty and attractive and models 0:20:01.190 --> 0:20:03.470 what you want to do is a fashion photographer or 0:20:03.470 --> 0:20:07.700 a beauty photographer is capture that natural beauty. The approach 0:20:07.700 --> 0:20:13.340 to maintaining that natural proportion and beauty is usually shot 0:20:13.340 --> 0:20:17.210 on a long focal length lens at a distance, so 0:20:17.480 --> 0:20:21.050 so that those proportions and parts of the face are 0:20:21.050 --> 0:20:24.109 maintained naturally. As soon as you start bringing the camera 0:20:24.109 --> 0:20:27.919 in closer, the proportions of the face, particularly the nose 0:20:27.920 --> 0:20:31.160 being closer to the camera, will start to exaggerate and 0:20:31.160 --> 0:20:34.760 those proportions must be disrupted. So part of the problem 0:20:34.760 --> 0:20:38.600 with superimposition is that you don't often know the distance 0:20:38.720 --> 0:20:42.679 between when that CCTV camera you can go and measure it, 0:20:42.680 --> 0:20:46.129 which I've done at crime scenes as well. But then 0:20:46.130 --> 0:20:48.260 you have to try to reproduce it. And there's a 0:20:48.260 --> 0:20:52.820 lot of, um, difficulties with trying that approach. It's not impossible, 0:20:52.820 --> 0:20:55.820 but superimposition is is certainly not the way to go 0:20:55.820 --> 0:20:59.210 with facial identification. A side by side like for like 0:20:59.210 --> 0:21:02.359 is is the method if it's an infrared versus a 0:21:02.359 --> 0:21:07.340 color image, that could be problematic because of the differences. 0:21:07.340 --> 0:21:10.880 They should be roughly the same angle, normal font size 0:21:10.880 --> 0:21:13.940 is the perfect angle, of course, but often that's not 0:21:13.940 --> 0:21:15.650 the case in surveillance photography. 0:21:16.830 --> 0:21:22.430 What about ease in terms of uniqueness? How reliable is 0:21:22.430 --> 0:21:25.730 analysis of ease and comparison in photo? 0:21:26.060 --> 0:21:30.230 The term uniqueness is being a bit of a dirty 0:21:30.230 --> 0:21:35.449 word in forensics over the last few years, because we've 0:21:35.450 --> 0:21:40.430 always claimed that fingerprints are unique. But from a scientific position, 0:21:40.430 --> 0:21:41.480 we can't prove that. 0:21:41.480 --> 0:21:44.149 Unless you've fingerprinted every single person who's ever lived on 0:21:44.150 --> 0:21:45.199 the planet, you. 0:21:45.200 --> 0:21:49.760 Can model it. The US government tried to model. Model fingerprint? Statistically, 0:21:49.790 --> 0:21:53.750 DNA does it through likelihood ratios. You know, to say 0:21:53.750 --> 0:21:57.350 that 1 in 20,000,000 chance of someone having the same DNA. 0:21:57.590 --> 0:22:02.990 So it can be modeled statistically. But uniqueness is something 0:22:02.990 --> 0:22:06.410 that we don't like to sort of, uh, say, well, 0:22:06.410 --> 0:22:09.560 we can say that they, they match or they have 0:22:09.560 --> 0:22:14.929 similar characteristics, um, as to satisfy a match or whatever, 0:22:14.930 --> 0:22:19.040 but not the concepts of uniqueness. So like fingerprints, there's 0:22:19.040 --> 0:22:22.190 a claim that is, ah, unique. But we can't prove 0:22:22.190 --> 0:22:26.270 that scientifically. And, you know, when you're comparing side by 0:22:26.270 --> 0:22:29.600 side is which is what we can do, you can 0:22:29.600 --> 0:22:32.659 say that there are a lot of strong anatomical similarities 0:22:32.660 --> 0:22:35.180 between the two ears, and the likelihood of it being 0:22:35.180 --> 0:22:38.960 the same is fairly high, but we we don't use 0:22:38.960 --> 0:22:43.159 the word uniqueness anymore. Um, that's something we've, we've got 0:22:43.160 --> 0:22:48.020 over it took fingerprinting, uh, bureaus quite a, quite a 0:22:48.020 --> 0:22:50.750 while to go away from that. And I think they 0:22:50.900 --> 0:22:56.000 may still not believe that fingerprints can't be proven. You know, um, 0:22:56.000 --> 0:23:01.370 Michael Sachs, uh, us sociologists has written really well on this, uh, 0:23:01.369 --> 0:23:05.960 around the, the fallacy of uniqueness and that, you know, 0:23:05.960 --> 0:23:12.680 if we're using scientific methods within our forensic analysis, you 0:23:12.680 --> 0:23:17.060 can't then come up with this qualitative claim that every fingerprints, 0:23:17.060 --> 0:23:21.050 unique or every is unique unless you can prove it scientifically, 0:23:21.050 --> 0:23:23.209 you can't really make that claim. And that was the 0:23:23.210 --> 0:23:26.720 fallacy of individualization. If you can't prove it scientifically, don't 0:23:26.720 --> 0:23:27.260 use it. 0:23:41.780 --> 0:23:48.050 With all these variables light, angles, perspectives, distance. How on 0:23:48.050 --> 0:23:55.040 earth is anyone actually identified by a photograph? Beyond all 0:23:55.040 --> 0:23:56.180 reasonable doubt. 0:23:57.350 --> 0:23:59.780 It seems you know, as much as I mentioned Martin 0:23:59.780 --> 0:24:02.540 Stevenson's paper, the third forensics, it looked like, you know, 0:24:02.540 --> 0:24:05.899 back in 2014 that this was going to be a revelation, 0:24:05.930 --> 0:24:09.890 a new way of identifying people with the predominant amount 0:24:09.890 --> 0:24:13.550 of imagery in the community, cameras in the community. This 0:24:13.550 --> 0:24:16.129 is going to be the the next best thing since 0:24:16.130 --> 0:24:22.910 DNA and fingerprints, but it just hasn't eventuated. I think 0:24:22.910 --> 0:24:27.739 the facial recognition software and the tools that might be 0:24:27.740 --> 0:24:31.879 used in the private sector and the verification of identity, 0:24:31.880 --> 0:24:35.389 I think that's that's where facial recognition is going to 0:24:35.390 --> 0:24:39.380 really play a part. But I'm not sure whether in 0:24:39.380 --> 0:24:43.910 a forensic situation that this form of identification evidence is 0:24:43.910 --> 0:24:48.020 going to resonate like the, like the fingerprints and DNA. 0:24:48.109 --> 0:24:52.310 I know the AFP's facial team, the last time I 0:24:52.310 --> 0:24:54.950 spoke with them, they hadn't gone to court. But they're 0:24:54.950 --> 0:24:59.660 working very hard in developing standards and developing training for 0:24:59.660 --> 0:25:02.270 their people. They're doing a very good job. But I 0:25:02.270 --> 0:25:06.410 think our initial thinking that being able to identify from 0:25:06.410 --> 0:25:10.400 it because it is in some way a trace like 0:25:10.400 --> 0:25:13.670 fingerprints and DNA left at the scene. But it's not 0:25:13.670 --> 0:25:19.070 a real direct trace. It's a visual trace. So there 0:25:19.070 --> 0:25:23.480 is some ambiguities and or differences between, I think that 0:25:23.480 --> 0:25:28.310 biological trace of a fingerprint or DNA versus a biological 0:25:28.310 --> 0:25:33.080 trace from a camera. There are just some, some complexities 0:25:33.080 --> 0:25:35.180 that it's just not made it as easy as what 0:25:35.180 --> 0:25:38.300 we thought it was. You know, if we time traveled, uh, 0:25:38.300 --> 0:25:42.830 50 years, not even 50 years, probably 20 years, 25 0:25:42.830 --> 0:25:47.000 years and say to detect these right in the modern you, 0:25:47.030 --> 0:25:50.479 you're going to have this resource where every crime is 0:25:50.480 --> 0:25:53.150 going to be recorded by a camera. Do you think 0:25:53.150 --> 0:25:56.000 they'll need detectives? And okay, well, of course not, because 0:25:56.000 --> 0:26:00.290 there it is. But it hasn't eventuated. Um, identifying people 0:26:00.290 --> 0:26:01.280 with difficult. 0:26:01.280 --> 0:26:05.780 It sounds to me like photographs are actually circumstantial evidence 0:26:06.020 --> 0:26:11.389 as opposed to definitive evidence. And I would ask, and 0:26:11.390 --> 0:26:16.370 we actually giving defense teams more to pick holes in 0:26:16.730 --> 0:26:21.290 by bringing photography and CCTV into it. 0:26:21.560 --> 0:26:24.380 That's a very important element, because what I've seen over the, 0:26:24.530 --> 0:26:28.040 in my experience, is the misuse of photographs as well. 0:26:28.040 --> 0:26:31.490 And this is this is a problem. I remember being 0:26:31.490 --> 0:26:34.100 in a biometric conference and I was on invited on 0:26:34.100 --> 0:26:37.879 as a panel of experts and, um, a guy I 0:26:37.880 --> 0:26:40.220 have a lot of respect for. Richard Brooke from the 0:26:40.220 --> 0:26:44.149 FBI was on that panel as well. And the question 0:26:44.150 --> 0:26:47.240 was asked of Richard. Richard was using a lot of 0:26:47.240 --> 0:26:52.820 facial identification in, um, in the FBI. Uh, the famous 0:26:52.820 --> 0:26:56.270 case that he worked on was the Afghan girl. I 0:26:56.270 --> 0:26:59.780 don't know whether you're familiar with Afghan girl, with, uh, 0:26:59.780 --> 0:27:04.880 Stephen McCurry. Um, National Geographic, the green eyes. The green eyes. Yeah. 0:27:04.880 --> 0:27:08.120 The green eye. Beautiful Afghan. Beautiful. Yeah, yeah. Absolutely stunning. 0:27:08.359 --> 0:27:10.760 They went back a few years later and tried to 0:27:10.760 --> 0:27:15.770 identify it. And I believe the FBI certainly made the identification. 0:27:15.890 --> 0:27:20.240 But it was asked of the panel, particularly Richard, of um, 0:27:20.930 --> 0:27:24.200 so what happens if you can't identify them, you know, 0:27:24.200 --> 0:27:27.859 so you've got a mugshot and exemplar photographed a source 0:27:27.859 --> 0:27:31.730 that you have known, known and known identity. And you've 0:27:31.730 --> 0:27:35.150 got a, a still of a, of a face from 0:27:35.150 --> 0:27:40.219 CCTV or ATM or a mobile phone, but the experts 0:27:40.220 --> 0:27:43.220 just can't identify it because, you know, maybe the quality 0:27:43.220 --> 0:27:45.830 isn't there or just it just you just can't identify it. 0:27:45.830 --> 0:27:48.470 What do you do? And he said, that's simple. We 0:27:48.470 --> 0:27:50.869 just put the two photographs together, paste them on a 0:27:50.869 --> 0:27:55.160 board and give them to the jury. And I sort 0:27:55.160 --> 0:27:58.400 of went, what do you do? What I said, so 0:27:58.400 --> 0:28:03.230 here you are, the FBI, who have all these trained 0:28:03.230 --> 0:28:08.720 science scientists and forensic experts trained in facial identification. They 0:28:08.720 --> 0:28:11.540 can't make the identification. But what you're asking the jury 0:28:11.540 --> 0:28:14.210 is to make that identification for you, because you can't. 0:28:14.450 --> 0:28:18.650 That's outrageous. That's that's biasing the evidence in a, in 0:28:18.650 --> 0:28:20.540 a in a very big way. And a good friend 0:28:20.540 --> 0:28:24.139 of mine, Gary Edmund from the unionist South Wales, he's 0:28:24.140 --> 0:28:27.740 position that he I've heard him say quite often is 0:28:27.740 --> 0:28:31.940 that particularly with fingerprints fingers fingerprints do the same. Fingerprint 0:28:31.940 --> 0:28:34.340 experts will have the exemplar and the fingerprint and the 0:28:34.340 --> 0:28:37.730 points of identification. Give that to the jury and they 0:28:37.730 --> 0:28:40.700 will see, you know, ten or 20 or 15 points 0:28:40.700 --> 0:28:43.110 of identification. The fingerprints. I don't have too much of 0:28:43.110 --> 0:28:47.820 a problem with that. But. But Professor Edmond suggests that. Well, 0:28:47.820 --> 0:28:49.980 if you've got the expert there, you don't need to 0:28:49.980 --> 0:28:53.490 show them that photographic evidence. And I think he's got 0:28:53.490 --> 0:28:57.240 a really strong point there that whether showing that sort 0:28:57.240 --> 0:29:00.660 of visual aspect to the jury, does it better inform 0:29:00.660 --> 0:29:07.050 them or is it influencing them in a wrong way? Another, um, 0:29:07.050 --> 0:29:09.930 encounter I had, it was with the New South Wales Police. 0:29:09.930 --> 0:29:12.720 I was asked by the chief scientists at the time 0:29:12.720 --> 0:29:17.729 to do a review of the forensic forensic photography, a 0:29:17.730 --> 0:29:21.180 forensic imaging unit in the New South Wales Police. And 0:29:21.180 --> 0:29:26.220 I did that with another AFP senior AFP officer. And 0:29:26.610 --> 0:29:29.580 we were shown around the the unit and we went 0:29:29.580 --> 0:29:32.250 to this, uh, one room where they had a poster 0:29:32.910 --> 0:29:39.120 and there was a CCTV photograph of two people holding 0:29:39.120 --> 0:29:41.580 up another two people in a park. I think it 0:29:41.580 --> 0:29:46.230 was in King's Cross. So then they had, uh, so 0:29:46.350 --> 0:29:48.810 just try to visualise this poster. So this big image 0:29:48.810 --> 0:29:53.010 in the middle was the CCTV image of the incident, 0:29:53.430 --> 0:29:57.450 a still image. There was then close ups of each 0:29:57.450 --> 0:30:01.800 person's head on the left hand side of the right 0:30:01.800 --> 0:30:05.100 hand side. Sorry of the poster. Then on the other 0:30:05.100 --> 0:30:08.580 opposite side of the poster was the two mug shots 0:30:08.580 --> 0:30:12.490 of the people. And I said, well, so what's this 0:30:12.490 --> 0:30:16.150 poster for? They said, oh, it's for court. Uh, identification 0:30:16.150 --> 0:30:19.240 of the people. Oh, you say you've identified them? Oh, no, no, no, 0:30:19.240 --> 0:30:21.940 we can't do that. Um, but we show them these 0:30:21.940 --> 0:30:25.150 posters so that the jury can, you know, uh, correlate 0:30:25.330 --> 0:30:28.960 between the the two close ups of the faces. So 0:30:28.960 --> 0:30:33.730 this is an example of issues I've come across where 0:30:33.730 --> 0:30:38.200 it's treating the jury as, uh, pseudo experts by showing 0:30:38.200 --> 0:30:42.460 them images. And we all believe we're very visually literate. 0:30:42.460 --> 0:30:47.410 We're a community that embellishes images all in everything we do. Even, 0:30:47.410 --> 0:30:49.600 you know, we're doing that now. We're recording this on 0:30:49.600 --> 0:30:53.560 on a camera, this conversation. So everything we do just 0:30:53.560 --> 0:30:57.940 just about is recorded in some visual form. I have 0:30:57.940 --> 0:31:02.680 some concerns when photographs are misrepresented to a jury in 0:31:02.680 --> 0:31:07.300 a seemingly unbiased, unaltered way for them to make up 0:31:07.300 --> 0:31:11.230 their mind. So using photographs, I think can be quite dangerous. 0:31:11.230 --> 0:31:14.110 And there's no filter with this at the moment in 0:31:14.110 --> 0:31:19.180 the criminal justice system, the judges, uh, that I've come across, uh, 0:31:19.480 --> 0:31:22.870 believe that the more information you can give the jury, 0:31:22.870 --> 0:31:25.990 the better the jury are to make the right decision. 0:31:26.230 --> 0:31:29.860 I certainly agree with that. But whether images can help 0:31:29.860 --> 0:31:33.880 or hinder the truth is something that's a bit concerning 0:31:33.880 --> 0:31:36.670 when it comes to visual evidence in criminal cases. 0:31:37.480 --> 0:31:41.770 There was a case that involved Larry David's show, Curb 0:31:41.770 --> 0:31:46.480 Your Enthusiasm, and the role that that played in a 0:31:46.480 --> 0:31:49.330 criminal trial. Can you go through that for us? 0:31:49.660 --> 0:31:53.230 Sure. Um, I wrote about it in my PhD thesis 0:31:53.230 --> 0:31:56.380 many years ago, but it's, um, it's a bit of 0:31:56.380 --> 0:32:02.560 a reverse, uh, identification. There was a gang murder, and 0:32:02.560 --> 0:32:06.100 one of the witnesses to that gang murder was actually 0:32:06.100 --> 0:32:11.200 murdered herself. And the accused of the original murder trial. 0:32:11.230 --> 0:32:15.850 His brother was charged with her murder, and he claimed 0:32:15.850 --> 0:32:18.670 that it couldn't have been him because he was at 0:32:18.670 --> 0:32:22.810 the baseball with his daughter. And he produced the seating 0:32:22.810 --> 0:32:25.030 stubs to the police. And they said, well, you could 0:32:25.030 --> 0:32:28.720 have got this off anyone. They tried to look it 0:32:28.720 --> 0:32:33.190 up with the CCTV cameras of the stadium, and they 0:32:33.190 --> 0:32:35.830 saw where his seat was, and they tried to enlarge it, 0:32:35.830 --> 0:32:40.030 but it was just too it just didn't allow definition 0:32:40.030 --> 0:32:43.240 that they could identify him sitting there with his daughter. 0:32:43.600 --> 0:32:47.800 But for some, uh, strange reason, Larry David was shooting, 0:32:47.800 --> 0:32:51.910 I think, Curb Your Enthusiasm show. And there was the 0:32:51.910 --> 0:32:54.490 accused in the background, sitting in his seat. 0:32:54.610 --> 0:32:55.510 With a good quality. 0:32:55.510 --> 0:33:00.370 Film. Yeah, broadcast quality film. And they use that. And 0:33:00.370 --> 0:33:03.670 of course, they got, um, uh, statements off the producers 0:33:03.670 --> 0:33:05.500 of the time and the and the time and date 0:33:05.500 --> 0:33:09.670 stamps on the, on the video. Um, and his alibi 0:33:09.670 --> 0:33:12.520 then was confirmed that he was at the baseball through 0:33:12.670 --> 0:33:16.480 the happened chance of being in the background of Larry 0:33:16.480 --> 0:33:19.240 David's show. So quite a quite an interesting. 0:33:19.300 --> 0:33:19.930 High definition. 0:33:19.930 --> 0:33:22.810 In high definition. Yeah, yeah. And he was telling the truth. 0:33:22.810 --> 0:33:25.570 So it can actually be used to exonerate. 0:33:25.720 --> 0:33:28.870 It can be used to exonerate. Absolutely. Any forensic evidence 0:33:28.870 --> 0:33:33.190 is about the first of all, trying to exonerate innocent 0:33:33.190 --> 0:33:37.840 people and secondly, trying to, uh, to convict or use 0:33:37.840 --> 0:33:41.200 to in the conviction of perpetrators of crime. But the 0:33:41.200 --> 0:33:44.500 primary objective is always to try to eliminate first, not 0:33:44.500 --> 0:33:48.250 not identify. Being an academic in forensic science at Western 0:33:48.250 --> 0:33:51.970 Sydney University for several years, being one of the founding 0:33:51.970 --> 0:33:55.900 academics or the founding academic there. For that course, we 0:33:55.900 --> 0:33:59.950 built a crime scene facility with the New South Wales Police, 0:33:59.950 --> 0:34:03.820 where students would go through and actually process scenes of crime. 0:34:04.120 --> 0:34:09.219 Forensic photography is more of the specialisation, the the enhancement 0:34:09.219 --> 0:34:13.989 of the of the evidence. Uh, the understanding reading of 0:34:13.989 --> 0:34:19.660 how vision works and how it how it affects the interpretation. Um, 0:34:19.660 --> 0:34:23.049 I think that's more about forensic crime scene photography really 0:34:23.050 --> 0:34:27.850 is more about recording in a document style, um, in 0:34:27.850 --> 0:34:31.360 a vernacular kind of photography, how the scene was found, 0:34:31.360 --> 0:34:35.680 items of evidence, murder, weapons, shoe marks and so forth. 0:34:35.680 --> 0:34:38.410 There are particular ways that they need to be photographed, 0:34:38.410 --> 0:34:41.530 but what a crime scene officer does in the processing 0:34:41.530 --> 0:34:45.040 of a scene is document. And they would they would 0:34:45.040 --> 0:34:49.990 document in images, usually video or still photography. They would 0:34:49.989 --> 0:34:53.890 write contemporaneous notes on where they saw things and how 0:34:53.890 --> 0:34:57.370 they found it, and also sketching, which is another visual, 0:34:57.370 --> 0:35:01.029 cultural sort of aspect to understanding crime scenes. So crime 0:35:01.030 --> 0:35:03.670 scene investigators would do those, those three things. They would 0:35:03.670 --> 0:35:07.810 also collect the evidence, but say shoe marks and fingerprints, 0:35:07.810 --> 0:35:11.270 items that are going to be analyzed. By other forensic 0:35:11.270 --> 0:35:15.770 experts need to be taken in a particular way so 0:35:15.770 --> 0:35:21.230 that distortion and artifacts are eliminated. And that's where forensic 0:35:21.230 --> 0:35:26.299 photography training has to occur with crime scene investigators. So 0:35:26.300 --> 0:35:29.660 they do do some, I guess, forensic photography on site 0:35:29.660 --> 0:35:33.830 when it comes to recording, documenting those types of evidence. 0:35:34.250 --> 0:35:39.230 But in a whole it's more around just recording the 0:35:39.230 --> 0:35:44.600 site as it's found and the critical elements of the site. 0:35:44.870 --> 0:35:48.470 They sort of work in a photo narrative form more 0:35:48.469 --> 0:35:52.790 so than anything, and they would use filmic techniques. So 0:35:52.790 --> 0:35:55.310 for instance, if there was a shoe mark in blood, 0:35:55.430 --> 0:35:59.359 they would photograph it using a style of photography, but 0:35:59.360 --> 0:36:02.000 in a way that that a filmmaker would approach it. 0:36:02.000 --> 0:36:04.460 So you do what we call an establishing shot. So 0:36:04.460 --> 0:36:07.100 you do a wide shot to show the location of 0:36:07.100 --> 0:36:08.989 the shoe mark in blood, and then you do a 0:36:08.989 --> 0:36:11.810 close up shot, which would be the forensic kind of 0:36:11.810 --> 0:36:14.480 where you would put a scale down. You would make 0:36:14.480 --> 0:36:17.930 sure that it's perpendicular to the subject. You would use 0:36:17.930 --> 0:36:21.440 lighting to show that the, the maximum detail of the, 0:36:21.440 --> 0:36:24.530 of the shoe mark. So it's very much a filmic 0:36:24.530 --> 0:36:29.480 kind of visual narrative that crime scene operatives work on 0:36:29.480 --> 0:36:33.290 in combination with their contemporaneous notes. And they use tools 0:36:33.290 --> 0:36:37.640 like markers, evidence markers to show and relate the photographs 0:36:37.640 --> 0:36:40.850 in the sketch to the contemporaneous notes. That's all a 0:36:40.850 --> 0:36:44.810 way of communicating a bloody shoe mark. That could be 0:36:44.870 --> 0:36:48.890 a item I found in the the lounge room or whatever, 0:36:48.890 --> 0:36:53.660 but that marker that's placed next to the the shoe mark, 0:36:53.660 --> 0:36:57.380 that item of evidence is only there really, so that 0:36:57.380 --> 0:37:01.730 the contemporaneous notes and the sketches and the photographs all 0:37:01.730 --> 0:37:03.920 tie in in a communicative way. And you. 0:37:03.920 --> 0:37:05.540 Can get a sense of scale and you can. 0:37:05.540 --> 0:37:08.210 Get a sense of scale. So we use linear scales 0:37:08.210 --> 0:37:12.140 to do that. Um, and, and, and all sorts of things. Yeah. 0:37:12.320 --> 0:37:14.660 And now with drones, you know, you can get aerial 0:37:14.660 --> 0:37:18.440 shots as well. Um, New South Wales Police, we were 0:37:18.440 --> 0:37:22.160 very big on using photogrammetry as well, where they would 0:37:22.160 --> 0:37:27.440 go in and use stereo cameras and record it, particularly for, uh, 0:37:27.440 --> 0:37:33.170 motor vehicle accidents particularly, and then reproduce that accurately. So 0:37:33.170 --> 0:37:37.040 measurements can be taken from their line drawings very similar 0:37:37.040 --> 0:37:39.560 to an old plan of a building. It's they sort 0:37:39.560 --> 0:37:42.680 of they can draw scenes in that sort of line sketch, 0:37:42.680 --> 0:37:45.890 but it's to scale and things can be measured. 0:37:47.050 --> 0:37:52.089 We've actually seen lots of visual illusions, and there's been 0:37:52.090 --> 0:37:55.360 so much fighting on the internet about, for example, the 0:37:55.360 --> 0:37:58.480 dress there was blue and gold and whatever people would 0:37:58.480 --> 0:38:02.230 argue to the death that they were totally different colours 0:38:02.230 --> 0:38:04.570 from what someone else was seeing. Can you give us 0:38:04.570 --> 0:38:09.310 an example of another visual illusion that people can actually understand? 0:38:09.310 --> 0:38:13.960 How our minds can be tricked, even though we're convinced otherwise? 0:38:14.530 --> 0:38:18.580 Yeah, there's lots of examples, Cathy. Um, I give a 0:38:18.580 --> 0:38:22.720 lecture in forensic photography or forensic criminology subject that I 0:38:22.719 --> 0:38:26.140 teach here at the university on how to interpret images. 0:38:26.140 --> 0:38:28.600 And basically what I'm trying to show the students is 0:38:28.600 --> 0:38:33.910 the ambiguity, how ambiguous a lot of the the vision is. 0:38:33.940 --> 0:38:37.930 An example is a horse race, uh, photo finish. You 0:38:37.930 --> 0:38:40.480 you would believe that they're in different places of the track, 0:38:40.480 --> 0:38:42.940 but in actual fact, the way that it's recorded through 0:38:42.940 --> 0:38:45.940 a slot, the horses are at the same location, but 0:38:45.940 --> 0:38:49.750 at different time. So in this type of recording, the 0:38:49.750 --> 0:38:53.529 variation is the time, not the distance, but our interpretation 0:38:53.530 --> 0:38:56.500 is the distance because we see a horse in front 0:38:56.500 --> 0:38:59.890 and a horse behind that the horse in in the crossing. 0:38:59.890 --> 0:39:02.590 The line first is though is in front of the 0:39:02.590 --> 0:39:06.700 horse coming second. But in actual fact, the photo, the 0:39:06.700 --> 0:39:10.719 imagery is all taken at the same location. The difference? 0:39:10.719 --> 0:39:15.730 There is time. But another great example is a checkerboard 0:39:15.730 --> 0:39:21.430 example by, um, a researcher called Edward Allison, who uses 0:39:21.430 --> 0:39:27.340 a visual anomaly where there's a if you can vision 0:39:27.430 --> 0:39:30.340 a black and white checkerboard or a chessboard like, well, 0:39:30.340 --> 0:39:34.480 like we all know, and there's a green cylinder sitting 0:39:34.480 --> 0:39:39.609 on the, uh, checkerboard casting a shadow over the checkerboard. 0:39:39.610 --> 0:39:44.500 And there are two, uh, squares that Alderson identifies, square 0:39:44.500 --> 0:39:48.640 A and square B, and the question is, is, is 0:39:48.640 --> 0:39:52.089 the tone of A and B square A and B 0:39:52.120 --> 0:39:56.860 the same or different? Now, unlike the dress that has people, 0:39:56.860 --> 0:40:02.620 C perceive the colors differently in this experiment, everyone that 0:40:02.620 --> 0:40:07.210 I'm aware of sees the two squares as different tones, 0:40:07.210 --> 0:40:10.960 blatantly different tones, and they would believe and swear that 0:40:10.960 --> 0:40:14.530 they are different tones. But if you were to isolate 0:40:14.530 --> 0:40:19.870 everything else and just look at those two squares like 0:40:19.870 --> 0:40:22.239 I do in my class, you'll see that that they're 0:40:22.239 --> 0:40:26.259 exactly the same tone of gray. I've even use Photoshop 0:40:26.260 --> 0:40:30.339 and measured the A square and the B square, and 0:40:30.340 --> 0:40:35.950 I get the same RGB value, uh, in both. So empirically, 0:40:35.950 --> 0:40:39.670 I've also proven that it's the same tonal range, but 0:40:39.670 --> 0:40:42.190 if you see it, uh, you would believe that a 0:40:42.190 --> 0:40:47.770 square A and square B is absolutely different tonality. But 0:40:47.770 --> 0:40:51.219 the reality is it's the same. And I'll go back 0:40:51.219 --> 0:40:54.399 to the Errol Morris quote, believing is seeing. It's not 0:40:54.400 --> 0:40:57.490 the other way around. You believe that it's different, but 0:40:57.489 --> 0:40:59.020 the facts are they're the same. 0:40:59.620 --> 0:41:03.370 I think my brain is just absolutely mashed up now 0:41:03.370 --> 0:41:08.770 because I thought photographs were definitive, and now I'm actually 0:41:08.770 --> 0:41:12.250 thinking it's all an illusion. Thank you so much, Glenn, 0:41:12.250 --> 0:41:18.219 for actually getting us thinking about photographs and reliability and 0:41:18.219 --> 0:41:24.130 evidence and facial recognition versus identification. It is so complicated. 0:41:24.340 --> 0:41:27.250 Thanks very much for joining us and explaining all of 0:41:27.250 --> 0:41:28.060 that today. 0:41:28.090 --> 0:41:29.710 You most welcome, Kathy anytime. 0:41:37.930 --> 0:41:42.400 Crime Insider's Forensics is a listener. Original production. It's hosted 0:41:42.400 --> 0:41:45.670 by me, Catherine Fox, and is produced by Ed Gordon. 0:41:46.120 --> 0:41:48.759 Sound Design and imaging is by Link Kelly.